论文部分内容阅读
本文以浙江省为研究区,在筛选并提取主要地理因子、计算主要城市节点可达性并利用Kriging插值方法获取可达性扩散表面的基础上,通过空间叠加将可达性与地理因子进行有效叠合,并采用空间采样方法获取采样数据,构造训练集。最后利用J48决策树算法对训练集进行归纳学习,构建并分析决策树,从而解释可达性与地理因子之间的相关模式。
In this paper, Zhejiang Province as the research area, in screening and extracting the main geographical factors, calculate the reachability of major cities and use Kriging interpolation method to obtain reachability diffusion surface, based on the spatial overlap will be reachability and geographic factors effective Overlap, and use the sampling method to obtain the sampling data space to construct the training set. Finally, the J48 decision tree algorithm is used to summarize the training set, and the decision tree is constructed and analyzed to explain the correlation between reachability and geographical factors.